jpountz commented on PR #12782:
URL: https://github.com/apache/lucene/pull/12782#issuecomment-1817146436

   Thanks @easyice. I took some time to look into the benchmark and improve a 
few things, hopefully you don't mind. Here is the output of the benchmark on my 
machine now:
   
   ```
   Benchmark                                   (numBytesPerInt)  (size)   Mode  
Cnt   Score   Error   Units
   GroupVIntBenchmark.byteArrayReadGroupVInt                  1      64  thrpt  
  5  24.483 ± 0.345  ops/us
   GroupVIntBenchmark.byteArrayReadGroupVInt                  2      64  thrpt  
  5  23.346 ± 0.288  ops/us
   GroupVIntBenchmark.byteArrayReadGroupVInt                  3      64  thrpt  
  5  16.318 ± 0.062  ops/us
   GroupVIntBenchmark.byteArrayReadGroupVInt                  4      64  thrpt  
  5  24.748 ± 0.993  ops/us
   GroupVIntBenchmark.byteArrayReadVInt                       1      64  thrpt  
  5  17.767 ± 0.081  ops/us
   GroupVIntBenchmark.byteArrayReadVInt                       2      64  thrpt  
  5   7.256 ± 0.013  ops/us
   GroupVIntBenchmark.byteArrayReadVInt                       3      64  thrpt  
  5   5.546 ± 0.449  ops/us
   GroupVIntBenchmark.byteArrayReadVInt                       4      64  thrpt  
  5   4.475 ± 0.021  ops/us
   GroupVIntBenchmark.byteBufferReadGroupVInt                 1      64  thrpt  
  5  21.812 ± 0.485  ops/us
   GroupVIntBenchmark.byteBufferReadGroupVInt                 2      64  thrpt  
  5  20.623 ± 1.454  ops/us
   GroupVIntBenchmark.byteBufferReadGroupVInt                 3      64  thrpt  
  5  13.601 ± 0.299  ops/us
   GroupVIntBenchmark.byteBufferReadGroupVInt                 4      64  thrpt  
  5  22.649 ± 0.662  ops/us
   GroupVIntBenchmark.byteBufferReadVInt                      1      64  thrpt  
  5  22.147 ± 0.083  ops/us
   GroupVIntBenchmark.byteBufferReadVInt                      2      64  thrpt  
  5   8.072 ± 0.116  ops/us
   GroupVIntBenchmark.byteBufferReadVInt                      3      64  thrpt  
  5   4.554 ± 0.394  ops/us
   GroupVIntBenchmark.byteBufferReadVInt                      4      64  thrpt  
  5   4.145 ± 0.674  ops/us
   ```
   
   The benchmark used to read directly from the in-memory byte[] by calling 
`rewind()` , I changed that to force it to read directly from the directoly to 
make the comparison a bit more fair.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org

Reply via email to